Events tracked by Clinfo
Clinfo receives status information about the cluster events from the Cluster Manager. This information is accessible by the routines in the APIs. Clinfo tracks topology events, as the cluster passes through various states.
Some of the states Clinfo tracks include:
- Cluster state is up or down
- Cluster substate has become stable or unstable
- Application has come online or failed over to a backup node
- Network has failed
- Node is in the process of joining the cluster
- Node has completed joining the cluster
- Node is leaving the cluster (that is, the node has failed)
- Node has abandoned the cluster
- New primary Cluster Manager has been elected (optional event)
- Change of state for a cluster site (if configured).
A complete listing of events can be found in subsequent sections of this document or in the clinfo.h include file, which is compiled into your application.
Clinfo receives dynamic reconfiguration events but does not track them; that is, applications cannot register to receive notification of dynamic reconfiguration events. Clinfo sets the cluster substate to CLSS_RECONFIG when it receives dynamic reconfiguration events. Applications can obtain this information using the cl_getcluster routine. The events triggered by the dynamic reconfiguration, such as a node up or node down event, are visible to applications.